The broad long-term objective of CA41464 is to increase the signal-to- noise ratio (SNR) of magnetic resonance imaging (MRI) through development of improved radiofrequency (rf) and gradient coils. In this competitive renewal, this objective is focused on the following subjects: (i) 3 Tesla MRI, recognizing that equipment manufacturers are not meeting the coil needs of investigators who use high field scanners; (ii) functional MRI (fMRI), a field in which the applicant's group has a substantial record of achievement; (iii) rf surface coils in the particular context of local gradient and rf transmit coils, recognizing the special problems that arise in this geometry; (iv) shim coils and susceptibility effects; and (v) consideration not only of the signal but also of the noise, recognizing that the noise in an echo-planar imaging (EPI) time course is physiological in origin. There are two interactive categories of Specific Aims: the first is hardware-based and the second is applications-based. In the first category, it is proposed to design, construct, and evaluate a new local head gradient coil that is optimized for spatial resolution within the 1988 FDA field switching guidelines. Image quality will be further enhanced by incorporating new rf coils and local shim coils. In the second category, a model for fMRI contrast-to- noise ratio (CNR), called the constant CNR (CCNR) model, is proposed and tested. Motor and visual paradigms will be used as a function of spatial resolution, echo time (TE), and repetition time (TR) in single- shot partial k-space gradient recalled EPI pulse sequences. fMRI research is moving forward through the work of many groups in several parallel streams: (i) basic cognitive neuroscience, to understand how the normal brain works at a systems level; (ii) physiology of the microcirculation of the brain using the tools of fMRI; (iii) diagnosis and presurgical planning in a context of neurosurgery, particularly involving cancer and epilepsy; (iv) diagnosis and response to treatment of patients with neurophsychiatric or neurodegenerative disease. The research proposed here is in support of these four streams of fMRI research.